If your LSA lead dispute was denied, the first thing to fix is the mental model, because the process you think you are fighting no longer exists. Around July to August 2024, Google retired manual lead disputes and replaced them with a machine-learning auto-credit system paired with a "Rate this lead" survey. There is no longer a form where you make a case and a reviewer accepts or rejects it. Instead, leads are assessed automatically—commonly within about 72 hours—and credits post within about 30 days. So a "denial" almost always means the model did not classify the lead as creditable, not that a person weighed your argument and said no. Understanding that distinction changes what you should do next.
What actually happens now when you flag a lead
Under the current system, you rate the lead through the survey and Google's model decides whether it qualifies for a credit. The model looks for signals that the lead was never a real opportunity. If it agrees, you get an automatic credit; if it doesn't, you simply don't—there is no manual escalation to re-argue that single lead the way the old dispute flow allowed. Your input still matters, but it matters as data that feeds the model's assessment and your account's pattern over time, not as a debate you win case by case.
This is why chasing an individual denial is usually the wrong instinct. The high-value behavior is rating every lead consistently and accurately so the system has clean signal, keeping good records, and reducing the junk that produces denials in the first place.
Creditable vs. not creditable
The single most common reason a lead isn't credited is that it was a real customer you simply didn't win. Google credits leads that were never genuine opportunities, not leads you lost. Here is the practical dividing line.
| Generally creditable | Generally NOT creditable |
|---|---|
| Spam or obvious junk contacts | A job type you do serve where the customer chose someone else |
| Wrong number / misdial with no service intent | A geography you do serve (job-type or geo mismatch inside your area) |
| Not a real customer (bots, solicitations, tests) | Price shoppers who decided you were too expensive |
| Service or location outside your profile per Google policy | Customer who booked a competitor after contacting you |
Notice the trap in the middle: a job-type or geo mismatch where you actually serve is not creditable. If someone in your service area asks for work you offer, that is a real lead even if it wasn't a fit for that particular call—Google treats it as a legitimate opportunity you were shown. The mismatches Google will credit are the ones that fall outside your stated profile per policy, not the ones inside it that simply didn't convert.
Two verticals are excluded entirely
If you operate in healthcare or tax, credits are excluded for your vertical. No amount of accurate rating changes that—there is no credit path to recover against. For those advertisers, the entire economic focus has to shift to lead quality and conversion, because the credit lever isn't available at all.
The number that reframes everything: ~6–7%
Here is the fact that should reset your expectations. Third-party estimates put recoverable spend at only about 6 to 7 percent. Credit recovery is a margin lever, not a rescue. Even a flawless rating operation gets back a single-digit slice of spend—useful, worth capturing, but never the thing that turns a losing LSA program into a winning one.
Contrast that with the other well-known estimate: a large share of raw leads—around 45 percent by third-party estimates—are unbookable. The math is stark. You can perfectly recover the ~6–7% that qualifies for credit, or you can attack the ~45% of leads that never turn into jobs at the source. The second is where the real money is. Denials are annoying; unbookable volume is expensive.
What to actually do after a denial
Given all of that, here is the productive playbook once a lead isn't credited.
- Rate every lead, consistently and accurately. The survey is your input to the model. Sloppy or inconsistent rating gives the system noisy signal and weakens your account's pattern over time.
- Keep call records and notes. Recordings, timestamps, and short notes on what the caller wanted help you rate accurately and give you evidence if a genuine pattern of bad leads needs to be raised with support.
- Escalate patterns, not single leads. One denial isn't worth a support ticket. A repeated pattern—say, a stream of out-of-area or spam contacts—is worth surfacing to Google support, because that is a systemic issue the survey alone may not resolve.
- Reduce junk at the source. Tighten service categories and service area so you are shown for the jobs you actually want, review your schedule so you answer fast, and fix the targeting that lets unbookable leads through. This is the lever that dwarfs credit recovery.
- Verify current rules. Google's credit criteria, timelines, and survey mechanics change. Confirm specifics against current Google Local Services Ads Help rather than assuming last year's process still applies.
Setting expectations with yourself
A denial is not a mistake to overturn—it is usually the model telling you the lead was a real opportunity you didn't win, or one that falls outside the narrow set Google credits. Accepting that lets you stop spending energy on a debate that no longer has a venue and start spending it where returns are larger: rating cleanly so the automatic credits you do deserve come through, and shrinking the unbookable share that is quietly the biggest line item in your lead economics.
The takeaway: since Google replaced manual disputes with an ML auto-credit model and a "Rate this lead" survey around mid-2024, a denial almost always means the system didn't classify the lead as creditable rather than a reviewer rejecting your case. Creditable leads are the ones that were never real—spam, wrong numbers, non-customers, or requests outside your profile per policy—while real customers you didn't win, including job-type or geo mismatches inside your service area, are not. Healthcare and tax are excluded, and recoverable spend is only about 6–7%, so treat credit recovery as a margin lever and put your real effort into cutting the roughly 45% of leads that are unbookable at the source. Confirm the current rules with Google, since they evolve.
Frequently asked questions
Can I still dispute an LSA lead the old way?
No. Around July to August 2024 Google retired the manual lead-dispute process and replaced it with a machine-learning auto-credit model paired with a Rate this lead survey. There is no manual form to argue a decision through anymore. Leads are assessed automatically, commonly within about 72 hours, and credits post within about 30 days. So a denial usually means the model did not classify the lead as creditable rather than a person reviewing and rejecting your case.
What kinds of LSA leads actually qualify for a credit?
Generally creditable leads are ones that were never real opportunities: spam, an obvious wrong number, someone who is not a genuine customer, or a request for a service or a location outside your profile per Google policy. What is not creditable is a real customer you simply did not win: a job type or geography you do serve, price shoppers, or someone who chose a competitor. Healthcare and tax verticals are excluded from credits entirely. Exact criteria can change, so check current Google Local Services Ads Help.
How much LSA spend can credit recovery realistically get back?
Not much on its own. Third-party estimates put recoverable spend at roughly 6 to 7 percent, which makes credit recovery a margin lever rather than a rescue. The bigger opportunity is reducing junk at the source, since a large share of raw leads, estimated around 45 percent, are unbookable. Consistent, accurate use of the Rate this lead survey plus better targeting does more for your economics than chasing individual denials.